Cultural Attraction among birds

If we could rewind the tape of cultural history and play it again, would it look like it does today ? Well, this is exactly what Olga Feher and collaborators at the Ofer Tchernichovski lab have tried to find out about zebra finches songs. Their Nature article shows the results of two ingenious series of experiments in cultural transmission. Their work nicely illustrates the phenomenon of attraction in social learning dynamics- a phenomenon that may have more weight in the animal cultures than has been recognized until now. They also provide a rather precise quantification procedure for the study of a multigenerational cultural trait (song learning) under a controlled setting- certainly something that until now you can afford more easily with zebra finches than with human subjects.

Oscine songbirds usually learn the song they sing from their conspecifics. This of coure make them fit for use as model organisms for the study of certain dynamics of cultural transmission. In their experiment Feher's team establishes a series of starting from isolate founders that have not had contact with other birds during their period of song learning, and who produce a song that clusters around certain features typical of what can be called the ''isolate song'', as opposed to the "wild-type song". These differences arise mainly as the result of lack of tutoring and not as the result of social deprivation.

So basically, their zebra finches are put to play the game of Chinese whispers. The ornithologists implemented two experimental settings now used in bird song learning: cultural transmission chains in which pupils become at their time tutors for the next generation in one-to-one settings (something that comes originally from social psychology studies with human subjects); and an isolate colony that reproduces for various generations in a large sound chamber with first-generation males and females that had no previous tutoring (a methodology that you should not propose to your local ethics committee for your anthropology dissertation).

In both conditions what they obtain is a very interesting set of results that can be described as the progressive transformation of the ''isolate song'' into a "wild-type song", through the various steps in the chain of social transmission. Does it mean anything for the study of animal cultures or even human culture, something miles apart from oscine song learning? Certainly due to the great diversity of song birds and their ecological adaptations, had the animal model been another species maybe the results would have been a bit different. But still the procedures and results of this carefully done study give food for thought to the animal cultures debate, and may catch the attention of the curious anthropologist.

First, the study describes the reproduction of cultural traits by generations of imitators. And though it shows how the song spreads through some form of vocal imitation - the general form of the ''syllables'' is basically reproduced - it also highlights the transformation that the song undergoes as it is transmitted from one generation to the next. You can actually listen to the transition between the isolate song and the wild-type song. That transformation takes place through a series of consistent and cumulative alterations, and ends up producing songs that are astoundingly similar from one (separate) lineage to the next. Had imitation been perfect or unbiased the different lineages would have not converged towards a similar song different from the original. Yet the songs of pupils deviated consistently tending towards the cluster of features that is the wild type song. As the authors put it: « Without such constraints, innovations and copying errors should cause unbounded variation over multiple generations or geographical distance, contrary to observations ».

Much of the debate currently surrounding animal cultures has concerned the mechanisms that could produce stable cultural traits - and as a result, most of the discussions (on chimpanzee culture, for instance) has focused on the role of imitation in producing cultural variability. Yet, as Bennett Galef once remarked, it is sometimes useful to dissociate the explanation of the causes of the spread of a social behavior, from the causes of its persistence or stability in the population. Returning to our zebra finch example, concerning its stability the authors report that the wild-type song is strongly determined by constraints (possibly phonological and neurological). It might seem paradoxical to some that the cultural song is also, biologically speaking, the most constrained.

Of course the assessment of these transformations has been mainly done through more sophisticated principal component analysis. This figure represents three different measures of the songs produced in the different lineages, with the arrows originating at the tutors and pointing towards pupils (colours represent lineages):

The purple shade represents the populational description of the wild type song. Note that the phenomenon of convergence towards the wild type song from tutors to pupils is rather well established for the first two features of the song (spectral-frame features and duration of acoustic state), but not for the third one (that concerns the dimension of rhythm). I could not help but finding this image reminiscent of another figure published, this time in Dan Sperber's article, in a section dealing with one of the multiple dynamics of cultural evolution, namely, ''attraction'' dynamics. The arrows also represent different cultural generations and the more obscure areas are those towards which the transformations of the cultural elements tend because of biological, psychological or cultural constraints.

Could it be possible – now, like poets, inspired by birds - to devise analogous methods for quantifying the phenomenon of attraction this time in the realm of human culture?